USVL-220 is a novel watermelon [Citrullus lanatus (Thunb.) Matsum. & Nakai] breeding line containing the nuclear genome of watermelon cultivars (Citrullus lanatus var. lanatus) and cytoplasmic background (chloroplast and mitochondrial genomes) derived from the desert species Citrullus colocynthis (L.) Schrad. USVL-220 was developed at the USDA, ARS, U.S. Vegetable Laboratory, Charleston, SC. Development of USVL-220 began in 1999 with the greenhouse observations of F1 plants derived from reciprocal crosses between the U.S. PI 386015 (C. colocynthis) and the watermelon cultivars New Hampshire Midget (NHM), Extra Early Sugar Baby, Allsweet, or Charleston Gray (C. lanatus var. lanatus). In general, F1 plants derived from a cross in which the C. colocynthis (PI 386015) was the maternal parent produced a female flower for every two to three male flowers. Conversely, in the reciprocal cross, in which a watermelon cultivar was used as the maternal parent, the F1 plants produced a female flower for every four to six (NHM) or five to seven (‘Allsweet’ or ‘Charleston Gray’) male flowers (Levi et al., 2006). Like with most plant species, the chloroplast and mitochondria are inherited maternally in watermelon (Havey et al., 1998; Levi and Thomas, 2005). Several studies have indicated that interaction between nuclear and maternally inherited (chloroplast or mitochondrial) genes affect flower morphogenesis in plant species (Ehlers et al., 2005; Kheyr-Pour, 1980; Ross, 1978; Ross and Gregourius, 1985; Van der Hulst et al., 2004; Wade and McCauley, 2005). Diolez et al. (1986) indicated that mitochondria play a role in the biosynthesis of endogenous ethylene in plant tissues, whereas Salman et al. (2008) suggested that ethylene reduces the formation of female flowers in watermelon. A wide genetic diversity exists between watermelon cultivars (C. lanatus var. lanatus) and U.S. PIs of C. lanatus var. citroides and C. colocynthis, indicating that a large number of exotic alleles may have been excluded during many years of cultivation and selection for watermelon with desirable fruit qualities (Levi et al., 2001). Our primary objective here was to enhance watermelon cultivars (C. lanatus var. lanatus) with the exotic chloroplast and mitochondrial genomes of the desert watermelon C. colocynthis as has been shown in other important crop plants (Burke and Stewart, 2003; Stewart, 1990; Tao et al., 2004). Such breeding lines should be valuable in the development of cytoplasmic substitution lines, which have the nuclear genome of a watermelon cultivar but the cytoplasm of C. colocynthis versus that of cultivated watermelon (C. lanatus var. lanatus) and could be useful in examining if any nuclear–cytoplasmic gene interaction affects female flower production in watermelon. The C. colocynthis exists in the hot deserts of North Africa, the Middle East, and South and Central Asia. As a desert plant, C. colocynthis can tolerate drought, intense sun exposure, and high day and low night temperatures better than the cultivated watermelon (Althawadi and Grace, 1986; Schafferman et al., 1998). Notable differences exist between the chloroplast genome of C. lanatus var. lanatus and that of C. colocynthis (Dane and Lang, 2004). Introducing the chloroplast and mitochondrial genomes of C. colocynthis into watermelon cultivars may improve their ability to survive drought conditions. Maternal effect on drought tolerance has been indicated in other crop plants, including the pondweed Potamogeton anguillanus (Iida et al., 2007) or Moss Tortula ruralis (Oliver et al., 2010). Cytoplasmic substation lines containing the C. colocynthis cytoplasm versus that of cultivated watermelon should be useful for studying if the cytoplasm plays any role in drought tolerance.